1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
use crossbeam_channel::{Receiver, Sender};
use csv::Reader;
use std::collections::HashSet;
use std::hash::Hasher;
use std::io::{Read, Seek};
use xxhash_rust::xxh3::{xxh3_128, Xxh3};
use crate::csv::Csv;
use crate::csv_hasher::CsvHasherExt;
use crate::csv_parse_result::{
CsvByteRecordWithHash, CsvLeftRightParseResult, CsvParseResult, CsvParseResultLeft,
CsvParseResultRight, Position, RecordHash, RecordHashWithPosition,
};
impl<R> CsvParseResult<CsvLeftRightParseResult<R>, R> for CsvParseResultLeft<R> {
#[inline]
fn new(record_hash: R) -> Self {
Self {
csv_left_right_parse_result: CsvLeftRightParseResult::Left(record_hash),
}
}
#[inline]
fn into_payload(self) -> CsvLeftRightParseResult<R> {
self.csv_left_right_parse_result
}
}
impl<R> CsvParseResult<CsvLeftRightParseResult<R>, R> for CsvParseResultRight<R> {
#[inline]
fn new(record_hash: R) -> Self {
Self {
csv_left_right_parse_result: CsvLeftRightParseResult::Right(record_hash),
}
}
#[inline]
fn into_payload(self) -> CsvLeftRightParseResult<R> {
self.csv_left_right_parse_result
}
}
pub(crate) struct CsvParserHasherLinesSender<T> {
sender: Sender<T>,
sender_total_lines: Sender<u64>,
}
impl CsvParserHasherLinesSender<CsvLeftRightParseResult<RecordHashWithPosition>> {
pub fn new(
sender: Sender<CsvLeftRightParseResult<RecordHashWithPosition>>,
sender_total_lines: Sender<u64>,
) -> Self {
Self {
sender,
sender_total_lines,
}
}
pub fn parse_and_hash<
R: Read + Seek + Send,
T: CsvParseResult<CsvLeftRightParseResult<RecordHashWithPosition>, RecordHashWithPosition>,
>(
&mut self,
csv: Csv<R>,
primary_key_columns: &HashSet<usize>,
) -> csv::Result<csv::Reader<R>> {
let mut csv_reader: Reader<R> = csv.into_csv_reader();
let mut csv_record = csv::ByteRecord::new();
// read first record in order to get the number of fields
if csv_reader.read_byte_record(&mut csv_record)? {
let csv_record_first = std::mem::take(&mut csv_record);
let fields_as_key: Vec<_> = primary_key_columns.iter().copied().collect();
// TODO: maybe use this in order to only hash fields that are values and not act
// as primary keys. We should probably only do this, if primary key field indices are
// contiguous, because otherwise we will have multiple calls to our hashing function,
// which could hurt performance.
// let num_of_fields = csv_record_first.len();
// let fields_as_value: Vec<_> = (0..num_of_fields)
// .filter(|x| !primary_key_columns.contains(x))
// .collect();
let record = csv_record_first;
let key_fields_iter = fields_as_key.iter().filter_map(|k_idx| record.get(*k_idx));
if key_fields_iter.peekable().peek().is_some() {
let key = record.hash_key_fields(fields_as_key.as_slice());
// TODO: don't hash all of it -> exclude the key fields (see below)
let hash_record = record.hash_record();
let pos = record.position().expect("a record position");
self.sender
.send(
T::new(RecordHashWithPosition::new(
key,
hash_record,
Position::new(pos.byte(), pos.line()),
))
.into_payload(),
)
.unwrap();
let mut line = 2;
while csv_reader.read_byte_record(&mut csv_record)? {
let key = csv_record.hash_key_fields(fields_as_key.as_slice());
let hash_record = csv_record.hash_record();
{
let pos = csv_record.position().expect("a record position");
self.sender
.send(
T::new(RecordHashWithPosition::new(
key,
hash_record,
Position::new(pos.byte(), pos.line()),
))
.into_payload(),
)
.unwrap();
}
line += 1;
}
self.sender_total_lines.send(line).unwrap();
}
} else {
self.sender_total_lines.send(0).unwrap();
}
Ok(csv_reader)
}
}
pub(crate) struct CsvParserHasherSender<T> {
sender: Sender<T>,
}
impl CsvParserHasherSender<CsvLeftRightParseResult<CsvByteRecordWithHash>> {
pub fn new(sender: Sender<CsvLeftRightParseResult<CsvByteRecordWithHash>>) -> Self {
Self { sender }
}
pub fn parse_and_hash<
R: Read + Send,
T: CsvParseResult<CsvLeftRightParseResult<CsvByteRecordWithHash>, CsvByteRecordWithHash>,
>(
&mut self,
csv: Csv<R>,
primary_key_columns: &HashSet<usize>,
receiver_csv_recycle: Receiver<csv::ByteRecord>,
) {
let mut csv_reader: Reader<R> = csv.into_csv_reader();
let mut csv_record = csv::ByteRecord::new();
// read first record in order to get the number of fields
match csv_reader.read_byte_record(&mut csv_record) {
Ok(true) => {
let record = std::mem::take(&mut csv_record);
let fields_as_key: Vec<_> = primary_key_columns.iter().copied().collect();
// TODO: maybe use this in order to only hash fields that are values and not act
// as primary keys. We should probably only do this, if primary key field indices are
// contiguous, because otherwise we will have multiple calls to our hashing function,
// which could hurt performance.
// let num_of_fields = record.len();
// let fields_as_value: Vec<_> = (0..num_of_fields)
// .filter(|x| !primary_key_columns.contains(x))
// .collect();
let mut hasher = Xxh3::new();
let mut key_fields_iter = fields_as_key
.iter()
.filter_map(|k_idx| record.get(*k_idx))
.peekable();
if key_fields_iter.peek().is_some() {
// TODO: try to do it with as few calls to `write` as possible (see below)
for key_field in key_fields_iter {
hasher.write(key_field);
}
let key = hasher.digest128();
// TODO: don't hash all of it -> exclude the key fields (see below)
let hash_record = xxh3_128(record.as_slice());
// we ignore any sending errors
let _ = self.sender.send(
T::new(CsvByteRecordWithHash::new(
Ok(record),
RecordHash::new(key, hash_record),
))
.into_payload(),
);
loop {
let mut csv_record = receiver_csv_recycle
.try_recv()
.unwrap_or_else(|_| csv::ByteRecord::new());
match csv_reader.read_byte_record(&mut csv_record) {
Ok(true) => {
hasher.reset();
let key_fields = fields_as_key
.iter()
.filter_map(|k_idx| csv_record.get(*k_idx));
// TODO: try to do it with as few calls to `write` as possible (see below)
for key_field in key_fields {
hasher.write(key_field);
}
let key = hasher.digest128();
// TODO: don't hash all of it -> exclude the key fields
// in order to still be efficient and do as few `write` calls as possible
// consider using `csv_record.range(...)` method
let hash_record = xxh3_128(csv_record.as_slice());
if self
.sender
.send(
T::new(CsvByteRecordWithHash::new(
Ok(csv_record),
RecordHash::new(key, hash_record),
))
.into_payload(),
)
.is_err()
{
// when the receiver is gone, it doesn't make sense to continue here
break;
}
}
Ok(false) => break,
Err(e) => {
if self
.sender
.send(
T::new(CsvByteRecordWithHash::new(
Err(e),
RecordHash::new(0, 0),
))
.into_payload(),
)
.is_err()
{
// when the receiver is gone, it doesn't make sense to continue here
break;
}
break;
}
}
}
}
}
Ok(false) => { /* Do nothing, we have reached EOF */ }
Err(e) => self
.sender
.send(
T::new(CsvByteRecordWithHash::new(Err(e), RecordHash::new(0, 0)))
.into_payload(),
)
.unwrap(),
}
}
}
#[derive(Debug)]
pub(crate) enum HashMapValue<T, TEq = T> {
Initial(u128, T),
Equal(TEq, TEq),
Modified(T, T),
}